[arin-ppml] RIPE/ITU

Milton L Mueller mueller at syr.edu
Fri Mar 5 05:55:08 EST 2010


>This made me realize that what we in the Internet community consider as
>an open participatory process, may not actually be open to everyone.  As
>technical people our organizations generally allow us a great deal of
>latitude to express our individual opinions.  This is generally not the
>case for government bureaucrats, especially in the realm of
>international diplomacy.  I'm not sure what to do about this, but we
>probably shouldn't just ignore it.

good observation. the light begins to dawn! I'm not sure what to do about it either, but I will say that ICANN's solution to this problem - the creation of a Govermental Advisory Committee with "special" powers over "public policy" - is NOT a good idea. this segregates governments into a silo and makes them advocates for governmental powers per se. In these discussions within ICANN, we have repeatedly pointed out that governmental representatives participate effectively in IETF, so a true bottom up model is feasible. But they are dealing with more technical issues than policy issues. for the bottom up model to work it takes a major cultural and institutional change in governments.

>Whereas, if large blocks of IPv6 addresses begin to
>be reserved for various constituencies such a scarcity is more likely to
>be created.  If this constituency receives a reservation, other equally
>deserving constituencies will not be far behind.  Such administrative
>actions are far more likely to create a scarcity than any actual
>allocation activity.

I agree, and that's actually one of the main motivations why the second report ITU commissioned - the one I did - considered the topic of TABLs or transferable blocks. 
If indeed an initial reservation creates a misallocation, then a TABL system would allow blocks to be transferred easily to those who need them and away from those who don't. 
The TABL proposal was actually a quite convervative in that it proposed to leave "needs-based allocations" in place for most of the v6 space but created this little safety valve of a section of the space that could be traded. 

>CIRs would have to follow the RIR's policies.  If this is really the
>case, then NIRs seem like a more appropriate way to serve the intended
>purpose of the CIRs.  If CIRs are intended to have policies independent
>of the RIRs then the conclusions of the paper must be called into question.

This is a slippery area. How do RIRs manage to have different, more locally-suited  policies without causing the same problems? What is the difference between CIRs and NIRs? 

>implementation or operational cost of an IPv6 network.  While fees are
>always an issue, it seems hard to believe that even the complete
>elimination of all IPv6 RIR fees could significantly impact the cost
>structure of implementing IPv6 for an ISP.

It's not the fees per se, it's the policies, which impose costs and barriers that, as you suggest, may be far more important than the annual fee. The idea is that if RIRs develop policies that are needlessly restrictive an alternative IR would try something different. 


More information about the ARIN-PPML mailing list